Atrial fibrillation driver identification through regional mutual information networks: a modeling perspective

Area of Science:

• Computational Biology
• Cardiac Electrophysiology
• Medical Imaging

Background:

• Atrial fibrillation ablation requires precise identification of electrical drivers.
• Tissue remodeling in atrial fibrillation complicates the detection of these drivers.
• Current mapping techniques may lack the resolution to differentiate complex activation patterns.

Purpose of the Study:

• To develop a mutual information, graph-based approach for identifying electrical drivers in remodeled atrial tissue.
• To propose local efficiency as a fault tolerance metric distinguishing rotational activation.
• To enhance the effectiveness of ablation treatments for atrial fibrillation.

Main Methods:

• Voltage data from 2D and 3D atrial tissue simulations were analyzed.
• Multi-spline catheter geometries were used for regional mapping.
• Graphs were constructed based on mutual information thresholds, and local efficiency was calculated.

Main Results:

• The derivative of local efficiency effectively distinguished rotational from irregular activation patterns across models.
• Rotational activations showed a decreased derivative of local efficiency compared to irregular activations (p < 0.01).
• The metric also differentiated varying degrees of atrial remodeling, indicating its sensitivity to tissue changes.

Conclusions:

• A decreased derivative of local efficiency is a characteristic of rotational activation in atrial tissue.
• This metric's variation with remodeling suggests its utility in assessing tissue state.
• The findings support the use of high-resolution mapping and this metric for identifying electrical drivers for targeted ablation.